Polygenic Transcriptome Risk Scores Improve Cross-Ancestry Portability of Risk Prediction for Pulmonary Function in the NHLBI Trans-Omics for Precision Medicine Program

• Activity Number: 70 - Novel Approaches for Omics and Multi-Omics Analysis
• Type: Contributed
• Date/Time: Sunday, August 7, 2022 : 4:00 PM to 5:50 PM
• Sponsor: Section on Statistics in Genomics and Genetics
• Author(s): Xiaowei Hu* and Dandi Qiao and Wonji Kim and Matthew Moll and Pallavi P Balte and Leslie A Lange and Traci M Bartz and Rajesh Kumar and Xingnan Li and Bing Yu and Brian E Cade and Cecelia A Laurie and Tamar Sofer and Ingo Ruczinski and Deborah A Nickerson and Donna M Muzny and Ginger A Metcalf and Harshavardhan Doddapaneni and Stacy Gabriel and Namrata Gupta and Shannon Dugan-Perez and L Adrienne Cupples and Laura R Loehr and Deepti Jain and Jerome I. Rotter and James G Wilson and Bruce M Psaty and Myriam Fornage and Alanna C Morrison and Vasan S Ramachandran and George Washko and Stephen S Rich and George T O’Connor and Eugene Bleecker and Robert C Kaplan and Ravi Kalhan and Susan Redline and Sina A Gharib and Deborah Meyers and Victor Ortega and Josée Dupuis and Stephanie J London and Tuuli Lappalainen and Elizabeth C Oelsner and Edwin K Silverman and R Graham Barr and Timothy A Thornton and Heather E Wheeler and Michael H Cho and Hae Kyung Im and Ani Manichaikul
• Companies: University of Virginia and Channing Division of Network Medicine, Brigham and Women's Hospital and Channing Division of Network Medicine, Brigham and Women's Hospital and Channing Division of Network Medicine, Brigham and Women's Hospital and Columbia University and School of Medicine Anschutz Medical Campus, University of Colorado and University of Washington and Division of Allergy and Clinical Immunology, Ann and Robert H. Lurie Children's Hospital and University of Arizona and Human Genetics Center, The University of Texas Health Science Center at Houston and Harvard Medical School and University of Washington and Harvard Medical School and Johns Hopkins Bloomberg School of Public Health and University of Washington and The Human Genome Sequencing Center, Baylor College of Medicine and The Human Genome Sequencing Center, Baylor College of Medicine and The Human Genome Sequencing Center, Baylor College of Medicine and Broad Institute of MIT and Harvard and Broad Institute of MIT and Harvard and The Human Genome Sequencing Center, Baylor College of Medicine and Boston University School of Public Health and University of North Carolina and University of Washington and The Lundquist Institute for Biomedical Innovation at Harbor-UCLA Medical Center and Division of Cardiovascular Medicine, Beth Israel Deaconess Medical Center and Cardiovascular Health Research Unit, University of Washington and Human Genetics Center, The University of Texas Health Science Center at Houston and Human Genetics Center, The University of Texas Health Science Center at Houston and Boston University and the National Heart Lung and Blood Institute’s Framingham Heart Study and Division of Pulmonary and Critical Care Medicine, Brigham and Women's Hospital and Center for Public Health Genomics, University of Virginia and Boston University and University of Arizona and Albert Einstein College of Medicine and Feinberg School of Medicine, Northwestern University and Harvard Medical School and Division of Pulmonary, Critical Care and Sleep Medicine, University of Washington and University of Arizona and Pulmonary and Critical Care, School of Medicine, Wake Forest University and Boston University School of Public Health and National Institutes of Health and New York Genome Center and Columbia University and Channing Division of Network Medicine, Brigham and Women's Hospital and Columbia University and University of Washington and Loyola University Chicago and Channing Division of Network Medicine, Brigham and Women's Hospital and Section of Genetic Medicine, The University of Chicago and University of Virginia
• Keywords: Polygenic Transcriptome Risk Score; Disease Risk Prediction; Cross-ancestry Portability; Integrative Analysis with Gene Expression; Pulmonary Disease

Abstract

While polygenic risk scores (PRS) enable early identification of genetic risk for Chronic Obstructive Pulmonary Disease (COPD), predictive performance is poor when the discovery and target populations are not well matched. Hypothesizing that the biological mechanisms of disease are shared across ancestry groups, we introduce a polygenic transcriptome risk score (PTRS) to improve cross-ancestry portability of risk prediction. We constructed the PTRS using summary statistics of a large-scale genome-wide association study of lung function in the UK Biobank. We examined prediction performance and cross-ancestry portability of PTRS by performing smoking-stratified analyses on about 40,000 multi-ancestry participants from the NHLBI Trans-Omics for Precision Medicine (TOPMed) program. While the PTRS had weaker association with COPD than PRS for European ancestry, the PTRS had stronger association than PRS for African Americans. Cross-ancestry portability of PTRS was significantly higher than PRS for COPD and across all smoking strata (paired t test with p< 2.2e-16). Our study demonstrates the value of PTRS compared to PRS for improved cross-ancestry portability in predicting COPD risk.